Development of a Common Database for Digital Mammography Research

Abstract

The purpose of this infrastructure project is to develop a large database of digitized mammograms that will be distributed free of charge to researchers working in all aspects of digital mammography. This database will facilitate and promote rapid development in digital mammography research. The database will consist of 1000 cases subdivided into 5 categories, 4 containing different breast lesions -- masses, microcalcifications, architectural distortions, asymmetric densities (both benign and malignant) -- and one containing normal mammograms. The mammograms will be collected and digitized (0.05-mm pixel size) at two sites: the Universities of Chicago and North Carolina. The database will be stored at the two sites and will be available over internet, and by mail on CD, tape, and magneto-optical disks. To date 630 cases have been digitized. Each case consists of index and previous exams (each having four standard views) and up to two special-view mammograms (e.g., magnification views). The computer systems for the database have been assembled and are connected to the network. The first release of 50 cases with clustered microcalcifications will be made shortly after true location of the cluster is determined. This release will be followed by another 50 cases of microcalcifications and 100 cases with masses. Along with the images, instructions on how to use the database will also be distributed. This includes an objective scoring method that we are developing based on input from radiologists'.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 2000
Accession Number
ADA391575

Entities

People

  • Robert Nishikawa

Organizations

  • University of Chicago

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Biomedical Research
  • Breast Cancer
  • Computer-Aided Diagnosis
  • Computers
  • Computing Devices
  • Continents
  • Databases
  • Distortion
  • Electronic Mail
  • Image Processing
  • Instructions
  • Internet
  • Mammography
  • Networks
  • Scientific Research
  • Standards
  • Universities

Fields of Study

  • Medicine
  • Physics

Readers

  • Computer Science/Computer Engineering/Data Science/Digital Signal Processing.
  • Database Systems and Applications
  • Oncology and Biomarker-Based Cancer Detection.